본문으로 바로 가기

How to use

The easiest way to go from daily sales to financial statements

From data prep to the confirm steps for accurate analysis

1. Getting started

Upload → quick confirm → report. Most fields are pre-filled — usually just a few clicks.

Upload a file

Drop a spreadsheet or JSON and we detect structure and headers automatically.

Quick confirm

Review industry, country, currency, and only when needed commercial area or column roles.

AI analysis & report

Get anomalies, seasonality, and benchmarks — then download PDF.

2. Prepare your data for better analysis

Common principles

  • Include a date (or month) column for trends, seasonality, and anomaly detection.
  • One header row is most reliable (rows 2–3 can still be detected).
  • Keep numeric columns as pure numbers — don’t mix unit text like “won” or “units”.
  • Total/subtotal rows are fine — we detect and separate them automatically.

Your uploaded data is never used to train AI models

Missing columns are OK. We analyze what you have and quietly skip metrics that can’t be computed.

Recommended columns by industry group

F&B (cafe, restaurant, delivery…)

Date, revenue, category/menu, channel (dine-in/delivery), transactions · nice to have: seats, delivery fee, delivery time

Retail / distribution

Date, revenue, category, channel/store · nice to have: inventory, shrinkage, fresh-food sales (grocery)

E-commerce

Date, GMV/revenue, orders, channel/category · nice to have: returns, ad spend/ROAS, AOV

SaaS / subscription

Date (month), MRR, customers, churn · nice to have: NRR, CAC, burn/runway

B2B services (IT consulting, agency…)

Date, revenue, project/client, utilization · nice to have: pipeline, billable hours, rate

Healthcare / wellness

Date, revenue, patients/members, new vs returning · nice to have: visit type, cancellation rate

Hospitality (hotel…)

Date, revenue, rooms/occupancy, ADR · nice to have: RevPAR, channel (OTA vs direct)

Education

Date, revenue, students, course · nice to have: new enrollments, churn, completion rate

Real estate / finance

Date, revenue/fees/returns, deal count · nice to have: vacancy, AUM, yield

Manufacturing / logistics / other

Date, revenue or production volume, defect rate, OEE · logistics: shipments/lead time · wholesale: accounts/inventory turns

3. Analysis flow

Each confirm step exists to improve accuracy. Most of the time it’s prefill + confirm.

  1. 1

    Industry

    KPIs, benchmarks, and actions depend on industry. Pick the closest match.

  2. 2

    Country · language · currency · fiscal year

    Sets report language, currency display, holidays, and fiscal calendar. No FX conversion — source currency stays as-is.

  3. 3

    Commercial area (F&B only)(conditional)

    Office vs tourist catchments change how we interpret spikes. Confirm the estimate from weekday/season patterns.

  4. 4

    Schema roles (uncertain columns only)(conditional)

    We ask only when stock vs flow (or similar) is ambiguous. Clear columns are auto-confirmed.

  5. 5

    Run analysis

    The server computes numbers; AI interprets and builds the report.

4. Supported file formats

Only these formats are accepted. UI copy and the file picker use the same list.

Excel(.xlsx/.xls/.xlsm), CSV/TSV, JSON